Is there a minimum item for questionnaire? Some researchers said that the total items for teenagers ranging from 80-110 items
As Penguin_Knight said in his comment, the issue of length has multiple facets. One of them is what's officially defined as "respondent burden", i.e., how easy or difficult it is for a respondent to complete the survey. The modern adolescent population is probably used to one-question surveys on Facebook and Google, so I don't see any way they can endure a 100-item instrument.
Typically, an assumed rate of how quickly people can go through a survey is about 3-4 questions per minute (in English on the phone; it may be faster or slower in other languages or in other modes -- web will be faster unless they really start reading rather than just clicking "Yes" in every column). At 100 items, you are looking at a 25-30 minute interview. Besides being 100 times more burdensome than what a teenager can sit through, you are at a point where you should start compensating for participation. This triggers an entirely different set of ethics problems of interacting with a protected population who is not yet fully mature and responsible for their actions (again, in the US, where you would have to pass the internal review boards to launch surveys on this population).
Welcome to the site.
No, there is no minimum number of items for a questionnaire.
Quite often, shorter questionnaires are better as they can get more responses.
Many many questionnaires have far less than 80 items.
And, finally what is "total items for teenagers" and who are these researchers?
Keep in mind all a questionnaire is, is a measuring device. The reason it is not a bathroom scale, is because you are not trying to measure weight but instead some other characteristic of, likely a person. For example their opinion on a particularly topic.
If a questionnaire can accurately measure the trait of interest in 5 questions than 5 questions is sufficient. The measure of accuracy is entirely up to person conducting the study, if low accuracy is sufficient (for whatever seemingly odd reason) then again just a few questions is likely enough.
Also remember, more questions doesn't necessarily correlate with higher accuracy. Where as the right questions are much more likely to measure the topic of interest effectively.